Orthogonal Matrices – Examples with Solutions
Explore the concepts of orthogonal matrices and their properties. Below you will find clear definitions alongside step-by-step examples and practice questions with detailed solutions.
Definition of Orthogonal Matrices
An \( n \times n \) matrix whose columns form an orthonormal set is called an orthogonal matrix.
As a reminder, a set of vectors is orthonormal if each vector is a unit vector (its length is \( 1 \)) and each vector is orthogonal to every other vector in the set.
Equivalently, \( Q \) is orthogonal if and only if \[ Q^{T} Q = Q\,Q^{T} = I_n. \]
These are examples of orthogonal matrices:
a) \[ \begin{bmatrix} 1 & 0 & 0 \\ 0 & 0 & 1 \\ 0 & -1 & 0 \end{bmatrix} \]
b) \[ \begin{bmatrix} \tfrac{1}{\sqrt{10}} & 0 & -\tfrac{3}{\sqrt{10}} \\ \tfrac{3}{\sqrt{10}} & 0 & \tfrac{1}{\sqrt{10}} \\ 0 & -1 & 0 \end{bmatrix} \]
c) \[ \begin{bmatrix} \tfrac{1}{\sqrt{6}} & 0 & \tfrac{5}{\sqrt{30}} \\ \tfrac{1}{\sqrt{6}} & -\tfrac{2}{\sqrt{5}} & -\tfrac{1}{\sqrt{30}} \\ \tfrac{2}{\sqrt{6}} & \tfrac{1}{\sqrt{5}} & -\tfrac{2}{\sqrt{30}} \end{bmatrix} \]
d) \[ \begin{bmatrix} \cos x & -\sin x \\ \sin x & \cos x \end{bmatrix} \]
Properties of Orthogonal Matrices
A list of the most important properties of orthogonal matrices is given below. If \( Q \) is an orthogonal matrix, then:
- \( Q^{-1} = Q^{T} \). This is the most important property: the inverse is simply the transpose.
- The rows of \( Q \) form an orthonormal set.
- \( Q^{-1} \) is also an orthogonal matrix.
- \( \det(Q) = \pm 1 \).
- If \( \lambda \) is an eigenvalue of \( Q \), then \( |\lambda| = 1 \).
- If \( Q_1 \) and \( Q_2 \) are \( n \times n \) orthogonal matrices, then the product \( Q_1 Q_2 \) is also orthogonal.
Examples with Solutions
Example 1 — Product of Orthogonal Matrices
The matrices \( Q_1 \) and \( Q_2 \) below are orthogonal. Verify that the product \( Q_1 Q_2 \) is also orthogonal (Property 6 above).
\[ Q_1 = \begin{bmatrix} 0 & 0 & 1 \\ 1 & 0 & 0 \\ 0 & 1 & 0 \end{bmatrix}, \quad Q_2 = \begin{bmatrix} 0 & -1 & 0 \\ 1 & 0 & 0 \\ 0 & 0 & -1 \end{bmatrix}. \]
View Solution to Example 1
First compute the product \( Q_1 Q_2 \):
\[ Q_1 Q_2 = \begin{bmatrix} 0 & 0 & 1 \\ 1 & 0 & 0 \\ 0 & 1 & 0 \end{bmatrix} \begin{bmatrix} 0 & -1 & 0 \\ 1 & 0 & 0 \\ 0 & 0 & -1 \end{bmatrix} = \begin{bmatrix} 0 & 0 & -1 \\ 0 & -1 & 0 \\ 1 & 0 & 0 \end{bmatrix}. \]
Let \( \mathbf{v}_1, \mathbf{v}_2, \mathbf{v}_3 \) be the columns of \( Q_1 Q_2 \):
\[ \mathbf{v}_1 = \begin{bmatrix} 0 \\ 0 \\ 1 \end{bmatrix}, \quad \mathbf{v}_2 = \begin{bmatrix} 0 \\ -1 \\ 0 \end{bmatrix}, \quad \mathbf{v}_3 = \begin{bmatrix} -1 \\ 0 \\ 0 \end{bmatrix}. \]
Each column has norm \( 1 \):
- \( \|\mathbf{v}_1\| = \sqrt{0^2 + 0^2 + 1^2} = 1 \)
- \( \|\mathbf{v}_2\| = \sqrt{0^2 + (-1)^2 + 0^2} = 1 \)
- \( \|\mathbf{v}_3\| = \sqrt{(-1)^2 + 0^2 + 0^2} = 1 \)
And any two distinct columns are orthogonal:
- \( \mathbf{v}_1 \cdot \mathbf{v}_2 = 0 + 0 + 0 = 0 \)
- \( \mathbf{v}_1 \cdot \mathbf{v}_3 = 0 + 0 + 0 = 0 \)
- \( \mathbf{v}_2 \cdot \mathbf{v}_3 = 0 + 0 + 0 = 0 \)
The columns of \( Q_1 Q_2 \) are orthonormal, so \( Q_1 Q_2 \) is orthogonal.
Example 2 — Inverse Equals Transpose
Use a calculator to find the inverse of the orthogonal matrix \( Q \) and verify Property 1 (\( Q^{-1} = Q^{T} \)).
\[ Q = \begin{bmatrix} 0 & 0 & 1 \\ -1 & 0 & 0 \\ 0 & -1 & 0 \end{bmatrix}. \]
View Solution to Example 2
Using a matrix calculator, we find the inverse:
\[ Q^{-1} = \begin{bmatrix} 0 & -1 & 0 \\ 0 & 0 & -1 \\ 1 & 0 & 0 \end{bmatrix}. \]
Now compute the transpose of \( Q \):
\[ Q^{T} = \begin{bmatrix} 0 & -1 & 0 \\ 0 & 0 & -1 \\ 1 & 0 & 0 \end{bmatrix}. \]
Since \( Q^{-1} = Q^{T} \), Property 1 is verified.
Example 3 — Find Constants for Orthogonality
Find the real constants \( a \) and \( b \) in the matrix \( Q \) such that \( Q \) is orthogonal.
\[ Q = \begin{bmatrix} -1 & 0 & 0 \\ 0 & \tfrac{1}{\sqrt{2}} & a \\ 0 & \tfrac{1}{\sqrt{2}} & b \end{bmatrix}. \]
View Solution to Example 3
Let \( \mathbf{v}_1, \mathbf{v}_2, \mathbf{v}_3 \) be the columns of \( Q \):
\[ \mathbf{v}_1 = \begin{bmatrix} -1 \\ 0 \\ 0 \end{bmatrix}, \quad \mathbf{v}_2 = \begin{bmatrix} 0 \\ \tfrac{1}{\sqrt{2}} \\ \tfrac{1}{\sqrt{2}} \end{bmatrix}, \quad \mathbf{v}_3 = \begin{bmatrix} 0 \\ a \\ b \end{bmatrix}. \]
Two conditions must be satisfied for \( Q \) to be orthogonal.
1) Each column has norm \( 1 \):
- \( \|\mathbf{v}_1\| = \sqrt{(-1)^2 + 0^2 + 0^2} = 1 \)
- \( \|\mathbf{v}_2\| = \sqrt{0 + \tfrac{1}{2} + \tfrac{1}{2}} = 1 \)
- \( \|\mathbf{v}_3\| = \sqrt{a^2 + b^2} = 1 \quad \text{(I)} \)
2) Any two distinct columns are orthogonal:
- \( \mathbf{v}_1 \cdot \mathbf{v}_2 = 0 \)
- \( \mathbf{v}_1 \cdot \mathbf{v}_3 = 0 \)
- \( \mathbf{v}_2 \cdot \mathbf{v}_3 = \tfrac{a}{\sqrt{2}} + \tfrac{b}{\sqrt{2}} = 0 \Rightarrow a + b = 0 \Rightarrow a = -b \quad \text{(II)} \)
Substituting \( a = -b \) into (I): \[ \sqrt{(-b)^2 + b^2} = \sqrt{2b^2} = 1 \;\Rightarrow\; b = \pm \tfrac{1}{\sqrt{2}}. \]
The two valid solutions are:
- \( a = -\tfrac{1}{\sqrt{2}} \) and \( b = \tfrac{1}{\sqrt{2}} \), or
- \( a = \tfrac{1}{\sqrt{2}} \) and \( b = -\tfrac{1}{\sqrt{2}} \).
Practice Questions
Part 1 — Verify Five Properties
Matrix \( Q \) is orthogonal. Verify Properties 1 through 5 for
\[ Q = \begin{bmatrix} 0 & 1 & 0 \\ -1 & 0 & 0 \\ 0 & 0 & 1 \end{bmatrix}. \]
View Solution to Part 1
Property 1: \( Q^{-1} = Q^{T} \).
\[ Q^{-1} = \begin{bmatrix} 0 & -1 & 0 \\ 1 & 0 & 0 \\ 0 & 0 & 1 \end{bmatrix}, \quad Q^{T} = \begin{bmatrix} 0 & -1 & 0 \\ 1 & 0 & 0 \\ 0 & 0 & 1 \end{bmatrix} \]
Hence \( Q^{-1} = Q^{T} \).
Property 2: the rows of \( Q \) form an orthonormal set.
The row vectors are \( \mathbf{v}_1 = (0, 1, 0)^T \), \( \mathbf{v}_2 = (-1, 0, 0)^T \), \( \mathbf{v}_3 = (0, 0, 1)^T \). Each has norm \( 1 \), and all pairwise inner products are \( 0 \).
Property 3: \( Q^{-1} \) is also orthogonal.
From Property 1, \( Q^{-1} = Q^{T} \), whose columns are \( \mathbf{v}_1, \mathbf{v}_2, \mathbf{v}_3 \), which we just verified form an orthonormal set. So \( Q^{-1} \) is orthogonal.
Property 4: \( \det(Q) = \pm 1 \).
\[ \det(Q) = -1 \cdot \det\begin{bmatrix} -1 & 0 \\ 0 & 1 \end{bmatrix} = -1 \cdot (-1) = 1. \]
Property 5: every eigenvalue satisfies \( |\lambda| = 1 \).
\[ \det(Q - \lambda I) = \det\begin{bmatrix} -\lambda & 1 & 0 \\ -1 & -\lambda & 0 \\ 0 & 0 & 1-\lambda \end{bmatrix} = (1 - \lambda)(\lambda^2 + 1) = 0. \]
So \( \lambda_1 = 1 \), \( \lambda_2 = i \), \( \lambda_3 = -i \), and each has magnitude \( 1 \).
Part 2 — Identify Orthogonal Matrices
Which of the following matrices are orthogonal?
\[ A = \begin{bmatrix} \tfrac{1}{\sqrt{2}} & \tfrac{1}{\sqrt{2}} \\ -\tfrac{1}{\sqrt{2}} & \tfrac{1}{\sqrt{2}} \end{bmatrix}, \quad B = \begin{bmatrix} -1 & 0 & 0 \\ 0 & 1 & 1 \\ 0 & 0 & 0 \end{bmatrix}, \quad C = \begin{bmatrix} \tfrac{1}{\sqrt{10}} & -\tfrac{3}{\sqrt{10}} & 0 \\ 0 & 0 & 1 \\ \tfrac{3}{\sqrt{10}} & \tfrac{1}{\sqrt{10}} & 0 \end{bmatrix}. \]
View Solution to Part 2
Matrices A and C are orthogonal.
Matrix B is not orthogonal: its columns 2 and 3 are not orthogonal (their dot product is \( 1 \)), and column 3 is the zero vector.
Part 3 — Find All Orthogonal Matrices
Find all matrices of the form \( A = \begin{bmatrix} p & q \\ \tfrac{1}{\sqrt{3}} & r \end{bmatrix} \) that are orthogonal.
View Solution to Part 3
Two conditions must hold for \( A \) to be orthogonal.
1) Both columns are unit vectors:
\[ \sqrt{p^2 + \tfrac{1}{3}} = 1 \quad \text{(I)}, \qquad \sqrt{q^2 + r^2} = 1 \quad \text{(II)}. \]
From (I): \( p^2 = \tfrac{2}{3} \), so \( p = \pm \sqrt{\tfrac{2}{3}} \).
2) The two columns are orthogonal:
\[ pq + \tfrac{r}{\sqrt{3}} = 0 \quad \text{(III)}. \]
Case \( p = \sqrt{2/3} \):
From (III): \( q = -\tfrac{\sqrt{2} r}{2} \). Substituting into (II): \[ \sqrt{\tfrac{2r^2}{4} + r^2} = \sqrt{\tfrac{3r^2}{2}} = 1 \;\Rightarrow\; r = \pm \sqrt{\tfrac{2}{3}}. \]
The corresponding pairs are \( (r = \sqrt{2/3}, q = -\sqrt{3}/3) \) and \( (r = -\sqrt{2/3}, q = \sqrt{3}/3) \).
Case \( p = -\sqrt{2/3} \):
Similarly, we get \( (r = \sqrt{2/3}, q = \sqrt{3}/3) \) and \( (r = -\sqrt{2/3}, q = -\sqrt{3}/3) \).
The four orthogonal matrices are:
\[ \begin{bmatrix} \sqrt{\tfrac{2}{3}} & -\tfrac{\sqrt{3}}{3} \\ \tfrac{1}{\sqrt{3}} & \sqrt{\tfrac{2}{3}} \end{bmatrix}, \quad \begin{bmatrix} \sqrt{\tfrac{2}{3}} & \tfrac{\sqrt{3}}{3} \\ \tfrac{1}{\sqrt{3}} & -\sqrt{\tfrac{2}{3}} \end{bmatrix}, \quad \begin{bmatrix} -\sqrt{\tfrac{2}{3}} & \tfrac{\sqrt{3}}{3} \\ \tfrac{1}{\sqrt{3}} & \sqrt{\tfrac{2}{3}} \end{bmatrix}, \quad \begin{bmatrix} -\sqrt{\tfrac{2}{3}} & -\tfrac{\sqrt{3}}{3} \\ \tfrac{1}{\sqrt{3}} & -\sqrt{\tfrac{2}{3}} \end{bmatrix}. \]
Part 4 — Inverse of an Orthogonal Matrix
Find the inverse of matrix \( A \):
\[ A = \begin{bmatrix} 0 & \tfrac{5}{3\sqrt{5}} & \tfrac{2}{3} \\ \tfrac{1}{\sqrt{5}} & -\tfrac{4}{3\sqrt{5}} & \tfrac{2}{3} \\ \tfrac{2}{\sqrt{5}} & \tfrac{2}{3\sqrt{5}} & -\tfrac{1}{3} \end{bmatrix}. \]
View Solution to Part 4
Instead of computing the inverse by row reduction, we first verify that \( A \) is orthogonal. The three columns \( \mathbf{v}_1, \mathbf{v}_2, \mathbf{v}_3 \) satisfy:
- \( \|\mathbf{v}_1\|^2 = 0 + \tfrac{1}{5} + \tfrac{4}{5} = 1 \)
- \( \|\mathbf{v}_2\|^2 = \tfrac{25}{45} + \tfrac{16}{45} + \tfrac{4}{45} = 1 \)
- \( \|\mathbf{v}_3\|^2 = \tfrac{4}{9} + \tfrac{4}{9} + \tfrac{1}{9} = 1 \)
and all pairwise dot products are zero:
- \( \mathbf{v}_1 \cdot \mathbf{v}_2 = 0 - \tfrac{4}{15} + \tfrac{4}{15} = 0 \)
- \( \mathbf{v}_1 \cdot \mathbf{v}_3 = 0 + \tfrac{2}{3\sqrt{5}} - \tfrac{2}{3\sqrt{5}} = 0 \)
- \( \mathbf{v}_2 \cdot \mathbf{v}_3 = \tfrac{10}{9\sqrt{5}} - \tfrac{8}{9\sqrt{5}} - \tfrac{2}{9\sqrt{5}} = 0 \)
So \( A \) is orthogonal and \( A^{-1} = A^{T} \):
\[ A^{-1} = \begin{bmatrix} 0 & \tfrac{1}{\sqrt{5}} & \tfrac{2}{\sqrt{5}} \\ \tfrac{5}{3\sqrt{5}} & -\tfrac{4}{3\sqrt{5}} & \tfrac{2}{3\sqrt{5}} \\ \tfrac{2}{3} & \tfrac{2}{3} & -\tfrac{1}{3} \end{bmatrix}. \]
More References and Links
- Matrices with Examples and Questions with Solutions
- Determinant of a Square Matrix
- Inverse Matrix Questions with Solutions
- Elementary Linear Algebra – 7th Edition – Howard Anton and Chris Rorres
- Introduction to Linear Algebra – Fifth Edition (2016) – Gilbert Strang
- Linear Algebra Done Right – Third Edition (2015) – Sheldon Axler
- Linear Algebra with Applications (2012) – Gareth Williams